So I see you're looking at Eran's problem.... ;-) Since you say he's fairly capable, I'm assuming when he said he had GC and MSLABS set up, he did it right, so a GC pause wouldn't cause the error.
Bad node? possible. It could easily be a networking/hardware issue which are pain in the ass problems to track down and solve. With respect to the dfs.bandwidthPerSec... yes its an HDFS setting. As you point out, its an indirect issue. However that doesn't mean it wouldn't have an impact on performance. OP states that this occurs under heavy writes. What happens to the writes when a table is splitting? On May 11, 2012, at 12:12 AM, Stack wrote: > On Thu, May 10, 2012 at 6:26 AM, Michael Segel > <michael_se...@hotmail.com> wrote:. >> 4) google dfs.balance.bandwidthPerSec I believe its also used by HBase when >> they need to move regions. > > Nah. This is an hdfs setting. HBase don't use it directly. > >> Speaking of which what happens when HBase decides to move a region? Does it >> make a copy on the new RS and then after its there, point to the new RS and >> then remove the old region? >> > > When one RS closes the region and another opens it, there is no copy > to be done since the region data is in the HDFS they both share. > > St.Ack >
